llama.cpp
Vulkan Bugfixes and Improvements
#7084
Merged

Vulkan Bugfixes and Improvements #7084

0cc4m merged 9 commits into master from 0cc4m/vulkan-improvements
0cc4m
0cc4m Modify mat mat mul shader for mul_mat_id, modify mat vec mul shaders …
2423c29c
0cc4m Merge remote-tracking branch 'origin/master' into 0cc4m/vulkan-moe
a124dfad
0cc4m Merge remote-tracking branch 'origin/master' into 0cc4m/vulkan-moe
1e46fa8d
0cc4m Further work towards MoE, disabled for now
3098206b
0cc4m Disable MoE code (not ready yet), fix a number of bugs in shaders and…
f3dc7040
0cc4m Add softmax with f16 mask and pos buffer support
6697303c
0cc4m Disable mul_mat_id shaders for now
947df475
0cc4m Fix flake8
d4abd657
MaggotHATE
0cc4m
daniandtheweb
MaggotHATE
0cc4m
MaggotHATE
MaggotHATE
teleprint-me
netrunnereve
ggerganov
ggerganov commented on 2024-05-08
ggerganov
ggerganov approved these changes on 2024-05-08
0cc4m
0cc4m
teleprint-me
teleprint-me
teleprint-me
0cc4m
slaren
0cc4m
mofosyne mofosyne added bugfix
mofosyne mofosyne added Review Complexity : High
0cc4m Fix validation errors caused by empty buffers on larger batch sizes
d2088cd7
0cc4m
MaggotHATE
0cc4m 0cc4m merged befddd0f into master 1 year ago
0cc4m 0cc4m deleted the 0cc4m/vulkan-improvements branch 1 year ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
Milestone